All eyes were on Silicon Valley and the Open Networking Summit this week. One of the big topics of conversation was Intel’s push into an already highly competitive software defined networking (SDN) space.
In theory, this move by Intel makes sense. SDNs transform the data center and create an opportunity for low-cost switch manufacturers to become a more important part of the data center. However, “theory” and “reality” are two different things, and I don’t believe a pure, white box switch really works in this market.
Over the past decade, chip companies such as Broadcom, Marvell, Mellanox and now Intel have tried to create software to complement the switch chips, in a bid to to sell the switch chips. Intel has WindRiver, Broadcom acquired LVL7, Marvell has Radian and even Mellanox uses open source software. Software plus chips equals success, right? Again, in theory it makes sense, but in practicality, this hasn’t worked.
In my opinion, chips are commoditizing and vendors are looking to differentiate more with software. However, if both software and hardware are provided by the same chip vendor, then the ODM is just a manufacturer and distributor. In this case, what is the difference to Cisco, Juniper and other traditional network vendors? At the end of the day, users want to have the best price performance for a network solution. The winning chip in the hardware may not necessarily have the best technical performance. It will be the chip vendor who knows how to work with independent software vendors.
This hasn’t worked because these companies do not and are not structured to provide customer support, services and an advanced operating system.
Looking back, this once played out in the server industry. When virtualization first arrived on the scene, we were saying the same things. The concept was that the ODMs would move into the market and take a chunk of share from the OEMs. The ODMs weren’t able to provide the necessary support so organizations continued to buy from the OEMs. I’ll agree that we’re starting to see this shift happen in the server space now, but it took 15 years for the industry to start moving that way. I believe network infrastructure is more difficult to build than servers, so this transition could easily take 20 years.
So, what’s a network manager to do if there’s a desire to move the SDN? Well, all isn’t lost, as there are a group of startups leveraging the white box switches to provide a solution that includes strong service and support around the operating system.
My recommendation would be to not look to the ODMs such as Quanta and Accton, but instead take a look at new SDN vendors such as Pica8, Plexxi and Pluribus. These vendors all tout the idea of leveraging white boxes and are acting more like appliance vendors, where their operating systems create the differentiation. And as a result, they don’t just sell switches but understand the challenges of SDNs and are set up to sell and then support their customers. White boxes work on paper, but they’ll leave companies hung out to dry when customers need the support.